kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-02-16 07:53:55 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

[https://nvbugs/5747920][fix] Fix multimodal serve test (#11296)

Browse Source 
Signed-off-by: yechank <161688079+yechank-nvidia@users.noreply.github.com>
This commit is contained in:
Yechan Kim 2026-02-05 15:12:53 +09:00 committed by GitHub
parent 8447a96c29
commit 36cb5f8c93
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
examples/serve/aiperf_client_for_multimodal.sh
View File

@ -2,7 +2,7 @@
aiperf profile \
-m Qwen2.5-VL-3B-Instruct \
--tokenizer Qwen/Qwen2.5-VL-3B-Instruct \
--tokenizer ${AIPERF_TOKENIZER_PATH:-Qwen/Qwen2.5-VL-3B-Instruct} \
--endpoint-type chat \
--random-seed 123 \
--image-width-mean 64 \

1
tests/integration/test_lists/waives.txt
View File

@ -215,7 +215,6 @@ full:sm89/accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_
full:sm89/accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[MMLU-tp2pp2] SKIP (https://nvbugs/5596337)
unittest/executor/test_rpc.py::TestRpcCorrectness::test_incremental_task_async SKIP (https://nvbugs/5741476)
test_e2e.py::test_trtllm_bench_llmapi_launch[pytorch_backend-llama-v3-llama3-8b] SKIP (https://nvbugs/5744432)
test_e2e.py::test_trtllm_serve_multimodal_example SKIP (https://nvbugs/5747920)
cpp/test_multi_gpu.py::TestDisagg::test_symmetric_executor[gpt-2proc-mpi_kvcache-90] SKIP (https://nvbugs/5755941)
examples/test_granite.py::test_llm_granite[granite-3.0-1b-a400m-instruct-bfloat16] SKIP (https://nvbugs/5608979)
examples/test_granite.py::test_llm_granite[granite-3.0-2b-instruct-bfloat16] SKIP (https://nvbugs/5608979)

7
tests/unittest/llmapi/apps/_test_trtllm_serve_multimodal_example.py
View File

@ -61,12 +61,15 @@ def example_root():
@pytest.mark.parametrize("exe, script",
[("python3", "openai_chat_client_for_multimodal.py"),
("bash", "aiperf_client_for_multimodal.sh")])
def test_trtllm_serve_examples(exe: str, script: str,
def test_trtllm_serve_examples(exe: str, script: str, model_name: str,
server: RemoteOpenAIServer, example_root: str):
client_script = os.path.join(example_root, script)
custom_env = os.environ.copy()
custom_env["AIPERF_TOKENIZER_PATH"] = get_model_path(model_name)
# CalledProcessError will be raised if any errors occur
subprocess.run([exe, client_script],
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
check=True)
check=True,
env=custom_env)